ChapsVision Quiet Roll-Up Now Holds French and German Intel Keys

Silvano Sansoni rejects the label outright. The managing director of ChapsVision told Tech.eu his firm is “a very different company” from Palantir after France’s DGSI picked the French data group to replace the US analytics giant. The contract, reported around €10m and covering terrorism-related data work, has pulled a once-discreet roll-up into the open.

What looks like a single sovereignty win sits on top of 29 acquisitions, a German intelligence deal already in hand, and talks with governments across northern Europe. That quieter build is the part most coverage skips. The political headline and the corporate construction are the same story viewed from opposite ends.

The Decade Contract France Finally Cut

Palantir first entered the DGSI after the 2015 Bataclan attacks. The 2016 deal was renewed in 2019, 2022 and again in December 2025. Six months later Prime Minister Sébastien Lecornu announced the switch.

1. 2015 Bataclan attacks open the door to foreign analytics tools
2. 2016 Palantir signs its first DGSI contract
3. 2019, 2022, December 2025 successive renewals keep the US platform in place
4. Late 2021 France opens a domestic-only tender that shuts Palantir out of the final race
5. Six months after the last renewal Lecornu announces the switch to ChapsVision

Lecornu framed it plainly: “We must use our own AI models; we cannot accept new strategic dependencies in the digital sphere. We cannot rely on tools developed by foreign powers. France must have its own tools.” The move landed amid wider European unease over US tech leverage, sharpened by recent model-access restrictions from American labs.

Sansoni, in the MD seat for one year, calls the DGSI the “front client” but says the scope runs across the entire French state. Migration of the complex existing systems will take 12 to 18 months. Some official statements have floated one to three years. Either way the hand-over is multi-year, not overnight.

• €10m reported value of the DGSI analytics package
• 12-18 months Sansoni’s migration window
• 2016-2025 Palantir’s decade of renewals now ending

Why the Sovereignty Clock Ran Out

European governments have watched US cloud and AI providers retain kill-switch potential. The Anthropic access curbs earlier this year revived the phrase “digital kill switch” in Brussels and Paris. France’s own tender for a domestic replacement had been open since late 2021 and limited to French firms, shutting Palantir out of the final race.

ChapsVision’s pitch lands on control. Customers keep the tech on their terms, with transparent stacks and no long lock-in. Sansoni says that demand comes from both governments and companies. The firm already lists Pfizer, Boeing and Exxon Mobil among commercial clients. Those names matter because they show the same control argument travels outside the intelligence world.

Germany moved first. In May the BfV domestic intelligence service selected ChapsVision’s platform over US rivals. A parliamentary oversight chair called it “a clear signal for European digital sovereignty.” The French decision now gives the same vendor two of Europe’s largest internal-security agencies.

The sequence itself carries weight. Berlin chose before Paris. The French award then turned a single national preference into a paired reference that sales teams can carry into every subsequent government conversation.

Twenty-Nine Deals Built the Quiet Contender

Founder Olivier Dellenbach sold eFront for more than $1.5bn in 2019 and launched ChapsVision the same year. The group has grown almost entirely by acquisition. Official tallies list 29 deals spanning CRM, AI search, phone tracing, translation and security.

The most visible recent step was the November 2024 purchase of Sinequa, a leader in enterprise search, paired with an €85m funding round. Earlier rounds brought in Bpifrance, Tikehau Capital, Qualium and Jolt Capital. Total capital raised sits around €275m to €350m depending on the count. The company now employs around 1,000 employees and nearly €200 million in 2024 revenue and sits in the French Tech Next40.

Integration remains the open task. Sansoni says the group aims to finish stitching the pieces together by 2027. Until then the platform story is still a collection of strong modules rather than one seamless product. That gap between acquired capability and unified delivery is the operational risk every buyer will probe.

The scale gap is real. Capability questions dominate the replies on X: Palantir bulls call ChapsVision “beta tech” and note that battlefield-hardened stacks take more than capital and political tailwinds. The firm’s answer is modernity. Younger code base, stronger security focus, and agentic tools such as translation already shipping.

What ArgonOS Delivers

The product at the centre is ArgonOS. ChapsVision describes the ArgonOS modular full-stack data platform as a way to fuse heterogeneous sources, surface weak signals, and keep everything on-premises, private cloud or air-gapped with zero external dependency.

• Unlimited connectors for structured, unstructured, sensor and OSINT feeds
• AI extraction from documents, images and media plus 60-plus language support
• Knowledge graphs, natural-language querying and RAG assistants
• Role-based access, full lineage and explainable outputs for audit
• Vertical foundations for defence, finance and compliance to cut deployment time

It is not a foundation-model vendor. It sits on top of customer data and supplies the decision layer. That distinction matters for agencies that already distrust external model providers. The architecture keeps the sensitive corpus inside the buyer’s perimeter while still offering the extraction, graph and assistant layers that operators now expect.

We are a very different company. I wouldn’t match ChapsVision as a European Palantir. For advertising it is a good comparison, but we are really a different company. We have different values. We work on a different philosophy.

Sansoni made the point repeatedly in the Tech.eu interview. The philosophy he cites is customer control and shorter lock-in versus the long, high-switching-cost relationships common in the sector.

Who Already Buys and Who Is Next

Customer mix today is still French-heavy: roughly 65 percent France, 20 percent United States, remainder across Europe, Japan and Singapore. Governments supply about 45 percent of revenue; the rest is commercial. Blue-chip names already on the books give the firm a reference base beyond intelligence.

Sansoni says the company is “currently talking with basically all of the European governments,” naming Poland, Germany, Denmark, Switzerland and Luxembourg. The German win is already public. The French contract expands the reference list further. Key growth markets he flags are the United States and Germany.

The same sovereignty pressure that opened the DGSI door is also creating space for other Palantir alumni and European challengers. One recent example is the ex-Palantir founders building a rival with fresh capital. Palantir’s own headquarters shift from Denver to Miami has done little to quiet European political concerns.

Commercial traction still matters for the narrative. Pfizer, Boeing and Exxon Mobil demonstrate that the control pitch works for regulated industries that face their own data-residency and audit pressures. Those logos travel into government meetings as proof the stack is not purpose-built only for spies.

Two Agencies Give the Sales Pitch Weight

Holding both the BfV and the DGSI on the same platform changes the conversation with every other European capital. A single national award can be dismissed as politics. Two of the continent’s largest internal-security services choosing the same vendor makes the technical and operational case harder to wave away.

Sansoni’s list of active talks already includes Poland, Denmark, Switzerland and Luxembourg alongside the German relationship. Each of those governments now has a concrete pair of references rather than a single French preference to evaluate. The pattern is no longer a one-off.

The commercial book supplies a second line of proof. When nearly half of revenue already comes from government work and the rest from large industrial names, buyers can see both sides of the house operating at once. That mix reduces the perception that the firm lives or dies on a single tender cycle.

The Integration Deadline Shapes Every Sale

The 2027 target for full integration of the 29 acquired pieces is now a public milestone. Until that work finishes, every new government deal arrives while the platform is still being stitched together. Buyers know they are purchasing a collection of strong modules that has not yet become one seamless product.

That fact cuts both ways. It keeps the story honest and forces ChapsVision to show progress on the seams rather than simply stacking more logos. It also means the DGSI migration window of 12 to 18 months will run in parallel with the broader integration push. Live operational pressure at French scale becomes an early stress test for the whole assembly.

If the seams hold, the dual FR-DE references become a durable sales tool. If they fray, sovereignty preference alone will not close the next round of tenders. The calendar therefore matters as much as the politics.

The Billion-Euro Target and 2030 IPO Clock

Sansoni’s stated goal is an IPO by 2030 at the latest, timed to roughly €1bn in revenue. That leaves an €800m gap. He expects 75 percent of the climb to come from further acquisitions and 25 percent from organic growth. Fresh funding rounds will finance the deal pipeline.

• €1bn revenue target timed to a 2030 IPO at the latest
• 75 percent of the climb expected from further acquisitions
• 25 percent expected from organic growth
• 2027 deadline to finish stitching the existing 29 deals

The firm already carries public-investment DNA through Bpifrance. That alignment helps when the buyer is a government ministry. It also raises the usual questions about speed of decision-making and commercial agility once the company is larger and more regulated.

Integration risk sits on the other side of the ledger. Twenty-nine deals create cultural and technical seams. The 2027 full-integration target is ambitious. If the seams hold, the dual FR-DE intelligence references become a powerful sales tool across the rest of Europe. If they fray, the sovereignty premium will not paper over product gaps.

For now the momentum is clear. A company that spent five years buying quietly has two of Europe’s most sensitive domestic agencies as customers and a pipeline of government conversations. The German BfV chose ArgonOS over Palantir months before Paris moved. The pattern is no longer a one-off French preference.

ChapsVision still has to prove the platform at DGSI scale under live operational pressure. The migration window will be the first real test. Success there turns a political story into a durable commercial one. Failure hands the next tender back to the incumbent or to the next French contender. Either outcome will be watched closely in Warsaw, Copenhagen and beyond.

Frequently Asked Questions

When was ChapsVision founded and by whom?

Olivier Dellenbach founded ChapsVision in May 2019 after selling eFront for more than $1.5 billion. He remains founder and chairman; Silvano Sansoni has served as managing director for the past year.

How many acquisitions has ChapsVision completed?

The company has completed 29 acquisitions to date, spanning search, CRM, translation, security and data tools. Full integration of those businesses is targeted for 2027.

What is the reported value of the DGSI contract?

The DGSI analytics package is reported at around €10 million. The scope extends beyond the intelligence service to other French state administrations and agencies.

What revenue target has ChapsVision set for its planned IPO?

Management aims for roughly €1 billion in revenue by a 2030 IPO at the latest, with 75 percent of the growth expected from further acquisitions and the rest organic.

How long will the migration from Palantir to ChapsVision take?

Sansoni estimates 12 to 18 months for the complex systems migration. Some government statements have described a window of one to three years while Palantir tools remain in parallel use.